• Journal of Korean Society for Geospatial Information Science
• /
• v.21 no.2
• /
• pp.19-25
• /
• 2013

Recently, the high-resolution satellite images is used the land cover and status data for the natural resources or environment management very helpful. The SVM algorithm of image processing has been used in various field. However, classification accuracy by SVM algorithm can be changed by various kernel functions and parameters. In this paper, the typical kernel function of the SVM algorithm was applied to the KOMPSAT-2 image and than the result of land cover performed the accuracy analysis using the checkpoint. Also, we carried out the analysis for selected the SVM kernel function from the land cover of the target region. As a result, the polynomial kernel function is demonstrated about the highest overall accuracy of classification. And that we know that the polynomial kernel and RBF kernel function is the best kernel function about each classification category accuracy.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.2 no.1
• /
• pp.32-44
• /
• 1999

Texture features can be incorporated in classification procedure to resolve class confusions. However, there have been few application-oriented studies made to evaluate the relative powers of texture analysis methods in a particular environment. This study evaluates the increases in the land-cover classification accuracy of the SPOT HRV multispectral data of Pusan Metropolitan area from texture processing. Twenty-four texture measures were derived from the SPOT HRV band 3 image. Each of these features were used in combination with the three spectral images in the classification of 10 land-cover classes. Supervised training and a Gaussian maximum likelihood classifier were used in the classification. It was found that while entropy produces the best empirical results in terms of the overall classification, other texture features can also largely improve the classification accuracies obtained by the use of the spectral images only. With the inclusion of texture, the classification for each category improves. Specially, urban built-up areas had much increase in accuracy. The results indicate that texture size 5 by 5 and 7 by 7 may be suitable at land cover classification of Pusan Metropolitan area.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.63 no.6
• /
• pp.27-37
• /
• 2021

Conversion of rice paddy field to upland has been accelerated as the central government incentivizes more profitable upland crop cultivation. The objective of this study was to investigate the current status and conversion trend from paddy to upland for the reservoir irrigation districts. Total 605 of reservoir irrigation districts whose beneficiary area is greater than 200 ha were selected for paddy-to-upland conversion analysis using the land cover maps provided by the EGIS of the Ministry of Environment. The land cover data of 2019 was used to analyze up-to-date upland conversion status and its correlation with city proximity, while land cover change between 2007 and 2019 was used for paddy-to-upland conversion trend analysis. Overall 14.8% of the entire study reservoir irrigation area was converted to upland cultivation including greenhouse and orchard areas. Approximately the portion of paddy area was reduced by 17.8% on average, while upland area was increased by 4.9% over the 12 years from 2007 to 2019. This conversion from paddy to upland cultivation was more pronounced in the Gyoenggi and Gyeongsang regions compared to other the Jeolla and Chungcheong provinces. The increase of upland area was also more notable in proximity of the major city. This study findings may assist to identify some hot reservoir districts of the rapid conversion to upland cultivation and thus plan to transition toward upland irrigation system.

• Journal of Ecology and Environment
• /
• v.42 no.4
• /
• pp.198-204
• /
• 2018

Background: The effective use of habitats is essential for the successful adaptation of a species to the local environment. Although habitats exhibit a hierarchical structure, including macro-, meso-, and microhabitats, the relationships among habitats of differing hierarchy have not been well studied. In this study, we studied the quantitative measures of microhabitat use of Gekko japonicus from three field populations in Japan: one at Tsushima Island, one at Nishi Park, Fukuoka, and one at Ohori Park, Fukuoka. We investigated whether land cover type, a higher hierarchical habitat component, was associated with quantitative microhabitat use, a lower hierarchical component, in these populations. Results: The substrate temperature where we located geckos (SubT) and the distance from the ground to the gecko (Height) were significantly different among the three populations. In particular, SubT on Tsushima Island was lower than it was in the other two populations. Irradiance at gecko location and Height were significantly different among the land cover types. In particular, Height in evergreen needleleaf forest was significantly lower than that in deciduous broadleaf forest. Furthermore, significant interactions between population and land cover type were observed for the SubT and Height variables. Conclusions: The quantitative measures of microhabitat use of G. japonicus varied with population and land cover type, which exhibited significant interaction effects on microhabitat use variables. These results suggest that higher hierarchical habitat components can affect the quantitative measures of lower hierarchical microhabitat use in nocturnal geckos.

• Korean Journal of Remote Sensing
• /
• v.35 no.2
• /
• pp.279-288
• /
• 2019

We evaluated the land cover classification performance of SegNet, which features semantic segmentation of aerial imagery. We selected four semantic classes, i.e., urban, farmland, forest, and water areas, and created 2,000 datasets using aerial images and land cover maps. The datasets were divided at a 8:2 ratio into training (1,600) and validation datasets (400); we evaluated validation accuracy after tuning the hyperparameters. SegNet performance was optimal at a batch size of five with 100,000 iterations. When 200 test datasets were subjected to semantic segmentation using the trained SegNet model, the accuracies were farmland 87.89%, forest 87.18%, water 83.66%, and urban regions 82.67%; the overall accuracy was 85.48%. Thus, deep learning-based semantic segmentation can be used to classify land cover.

• Korean Journal of Remote Sensing
• /
• v.38 no.6_4
• /
• pp.1911-1923
• /
• 2022

The availability of high-resolution satellite images provides precise information without direct observation of the research target. Korea Multi-Purpose Satellite (KOMPSAT), also known as the Arirang satellite, has been developed and utilized for earth observation. The machine learning model was continuously proven as a good classifier in classifying remotely sensed images. This study aimed to compare the performance of the support vector machine (SVM) model in classifying the land cover of the Delaware River port area on high and medium-resolution images. Three optical images, which are KOMPSAT-2, KOMPSAT-3A, and Sentinel-2B, were classified into six land cover classes, including water, road, vegetation, building, vacant, and shadow. The KOMPSAT images are provided by Korea Aerospace Research Institute (KARI), and the Sentinel-2B image was provided by the European Space Agency (ESA). The training samples were manually digitized for each land cover class and considered the reference image. The predicted images were compared to the actual data to obtain the accuracy assessment using a confusion matrix analysis. In addition, the time-consuming training and classifying were recorded to evaluate the model performance. The results showed that the KOMPSAT-3A image has the highest overall accuracy and followed by KOMPSAT-2 and Sentinel-2B results. On the contrary, the model took a long time to classify the higher-resolution image compared to the lower resolution. For that reason, we can conclude that the SVM model performed better in the higher resolution image with the consequence of the longer time-consuming training and classifying data. Thus, this finding might provide consideration for related researchers when selecting satellite imagery for effective and accurate image classification.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.3 no.4
• /
• pp.215-219
• /
• 2001

Land surface parameterization schemes for atmospheric models as well as decision support tools for ecosystem management require a frequent updating of land cover classification data for regional to global scales. Rice paddies have not been treated independently from other agricultural land classes in many classification systems, despite their atmospheric and ecological significance. A simple but improved method over conventional land cover classification schemes for rice paddy is suggested. Normalized difference vegetation index (NDVI) was calculated for the land area of South Korea at a 1km by 1 km resolution from the visible and the near-infrared channel reflectances of NOAA-AVHRR (Advanced Very High Resolution Radiometer). Monthly composite images of daily maximum NDVI were prepared for May and August, and used to classify 4 major land cover classes : urban, farmland, forests and water body. Among the pixels classified as "forests" in August, those classified as "water body" in May were assigned a "rice paddy" class. The distribution pattern of "rice paddy" pixels was very similar to the reported rice acreage of 1,455 Myons, which is the smallest administrative land unit in Korea. The correlation coefficient between the estimated and the reported acreage of Myons was 0.7, while 0.5 was calculated from the USGS classification.calculated from the USGS classification.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.6 no.3
• /
• pp.117-128
• /
• 2003

The purpose of this study is to analyze the characteristics of the landscape type on Chunchon by CG(computer graphics) pictures and visibility analysis. The land use CG picture and the land cover CG picture are created by using Zoning area data and DEM(digital elevation model), and by using data of land cover classification and DEM. According to the analysis result of the land cover from 1989 to 2000, the city area has increased to $7.7km^2$, the green area has diminished to $12.7km^2$. The tendency of the city area increases and the green area decreases which appear in the city area, developmental restriction zone and green area on land use. The landscape is classified into three types by cluster analysis using the area rate of the element which constitutes the land use CG picture. Type 1 is a landscape characteristics of developmental restriction zone. Type 2 is a landscape characteristics of green land and type 3 is a landscape characteristics based on city area and water area. The increase of city area and decrease of green area are shown in all landscape types of land cover CG pictures. The same tendency is seen in the place where the scenery is of high importance as a result of visible analysis. The preservation and management of the scenery to the green area are requird in developmental restriction zone are required.

• Journal of Cadastre & Land InformatiX
• /
• v.49 no.2
• /
• pp.151-167
• /
• 2019

Korea's current Land Category Classification System(LCCS) can not accurately register of complex and diverse Status of land use in Cadastral Record. Therefore, in order to draw implications for the improvement of LCCS in Korea, Shin SW and four others published a paper titled 'A Study on Land Category System of Domestic and Foreign Country' in 2013. This paper compared the 'land category', 'land use' and 'land cover' of six countries on the same line, and Some non-factual content was described. So, presented a new perspective on this. Looking forward, I hope that reasonable alternative will be presented based on the understanding of LCCS of Germany, Japan and Taiwan. In the future research project, to study the history of LCCS in Germany and Taiwan and suggest to refer to improvement of LCCS of Korea.

• Journal of Forest and Environmental Science
• /
• v.34 no.3
• /
• pp.246-248
• /
• 2018

Urban development and population have augmented the increase of impervious land-cover. This phenomenon has amplified the effects of climate change and increasing urban island effects due to increases in urban temperatures. Seoul, South Korea is one of the largest metropolitan cities in the world. While land uses in Seoul vary, land cover patterns have not changed much (under 2%) in the past 10 years, making the city a prime target for studying the effects of land cover types on the urban temperature. This research seeks to generalize the urban temperature of Seoul through a series of statistical tests using multi-temporal remote sensing data focusing on multiple scales and typologies of green space to determine its overall effectiveness in reducing the urban heat. The distribution of LST values was reduced as the size of urban forests increased. It means that changing temperature of large-scale green-spaces is less influenced because the broad distribution could be resulted in various external variables such as slope aspect, topographic height and density of planting areas, while small-scale urban forests are more affected from that. The large-scale green spaces contributed significantly to lowering urban temperature by showing a similar mean LST value. Both of concentration and dispersal of urban forests affected the reduction of urban temperature. Therefore, the findings of this research support that creating urban forests in an urban region could reduce urban temperature regardless of the scale.